Archives/gpt4all
2
0
Fork 0
mirror of https://github.com/nomic-ai/gpt4all synced 2024-11-10 01:10:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

llamamodel: fix embedding crash for >512 tokens after #2310 (#2383)

Browse Source 
Signed-off-by: Jared Van Bortel <jared@nomic.ai>
This commit is contained in:
Jared Van Bortel 2024-05-29 10:51:00 -04:00 committed by GitHub
parent f047f383d0
commit e94177ee9a
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
gpt4all-backend/llamamodel.cpp
View File

@ -386,7 +386,8 @@ bool LLamaModel::loadModel(const std::string &modelPath, int n_ctx, int ngl)
bool isEmbedding = is_embedding_arch(llama_model_arch(d_ptr->model)); bool isEmbedding = is_embedding_arch(llama_model_arch(d_ptr->model));
const int n_ctx_train = llama_n_ctx_train(d_ptr->model); const int n_ctx_train = llama_n_ctx_train(d_ptr->model);
if (isEmbedding) { if (isEmbedding) {
d_ptr->ctx_params.n_batch = n_ctx; d_ptr->ctx_params.n_batch = n_ctx;
d_ptr->ctx_params.n_ubatch = n_ctx;
} else { } else {
if (n_ctx > n_ctx_train) { if (n_ctx > n_ctx_train) {
std::cerr << "warning: model was trained on only " << n_ctx_train << " context tokens (" std::cerr << "warning: model was trained on only " << n_ctx_train << " context tokens ("